The question of how tropical trees cope with infertile soils has been challenging to address, in part, because fine root dynamics must be studied in situ. We used annual fertilization with nitrogen (N as urea, 12.5 g N m−2 year−1), phosphorus (P as superphosphate, 5 g P m−2 year−1) and potassium (K as KCl, 5 g K m−2 year−1) within 38 ha of old-growth lowland tropical moist forest in Panama and examined fine root dynamics with minirhizotron images. We expected that added P, above all, would (i) decrease fine root biomass but, (ii) have no impact on fine root turnover. Soil in the study area was moderately acidic (pH = 5.28), had moderate concentrations of exchangeable base cations (13.4 cmol kg−1), low concentrations of Bray-extractable phosphate (PO4 = 2.2 mg kg−1), and modest concentrations of KCl-extractable nitrate (NO3 = 5.0 mg kg−1) and KCl-extractable ammonium (NH4 = 15.5 mg kg−1). Added N increased concentrations of KCl-extractable NO3 and acidified the soil by one pH unit. Added P increased concentrations of Bray-extractable PO4 and P in the labile fraction. Concentrations of exchangeable K were elevated in K addition plots but reduced by N additions. Fine root dynamics responded to added K rather than added P. After 2 years, added K decreased fine root biomass from 330 to 275 g m−2. The turnover coefficient of fine roots <1 mm diameter ranged from 2.6 to 4.4 per year, and the largest values occurred in plots with added K. This study supported the view that biomass and dynamics of fine roots respond to soil nutrient availability in species-rich, lowland tropical moist forest. However, K rather than P elicited root responses. Fine roots smaller than 1 mm have a short lifetime (<140 days), and control of fine root production by nutrient availability in tropical forests deserves more study.